Framing data justice: An international development perspective.

Photo by Desola Lanre-Ologun on Unsplash

The field of international development informatics is growing rapidly. The World Bank plays a significant role in driving up the datafication of government services in the Global South. For example, the bank “has committed more than US $1.5 billion in financing to help 35 countries build digital identification systems and update their population records — a move the Bank says will help governments quickly verify the identities and needs of people seeking access to public services” (Human Rights Watch, 2023 p 26).

However, just as in the world’s high-income countries, such initiatives can bring harm and benefits, especially when implemented without due regard to the principles of data justice.

Richard Heeks and Satyarupa Shekhar, researchers in the field of international development informatics, created a framework for analysing data justice. They describe it as a conceptual model of data justice for international development studies, but one that applies to other domains.

Conceptual model of data justice (Heeks & Shekhar, 2019)

The model includes five dimensions of data justice defined in the following way:

• Procedural: fairness in the way in which data is handled–including the source of data, how data are captured, processed and used to make decisions.
• Instrumental: fairness in the results of data being used–including the outcomes and impact of data use, both intended and unintended.
• Rights-based: adherence to basic data rights such as representation, privacy, access and ownership.
• Structural: the degree to which the interests and power in wider society support fair outcomes in other forms of data justice–including the extent to which existing structural inequities in the context of the data initiative (see diagram above) are reduced or exacerbated.
• Distributive: an overarching dimension relating to the (in)equality of data-related outcomes that can be applied to each of the other dimensions of data justice.

In a series of case studies from different countries focusing on instances of digital community mapping of cities in the global South, Heeks and Shekhar (2019) use the model to highlight the benefits and disbenefits of the mapping exercise. The process of mapping urban areas, including informal settlements and slum areas, was undertaken “to counter the relative invisibility of (i.e., lack of data about) marginalised communities by gathering, visualising and utilising new data on locations, assets and issues within those communities”.

The article presents a full analysis of these initiatives against the five dimensions and I have extracted a few examples below to illustrate the process.

Considering the dimension of procedural justice, Heeks and Shekhar (2019) found that “these pro-equity data initiatives were somewhat ‘extractive’ in utilising a few community residents as data sources but largely excluding them from all other information value chain processes”.

Considering rights-based data justice, Heeks and Shekhar (2019) point out that these rights include the right to be represented in a dataset; but also the right to withhold data, the right to privacy, and to have some control over what and who is visible to whom, especially in the context of the grossly unequal power relations between government and NGO actors, and the people dwelling in slums.

In their discussion of the digital mapping of slum areas they highlight how making these geographic representations more visible challenged residents right to control their data:

These slum areas — all of which now exist virtually in some form of web-based map — are now legible not just to the state but to local and international NGOs, donor agencies, media organisations, academics, etc…All of these can know the slum to some extent and make decisions and actions using data about the slum without the permission or even the knowledge of anyone living there. The right to be represented can thus mean that the right to own and control is ceded to these external agents and to their particular interests and agendas, benevolent or otherwise. (p. 1001)

Indeed, during the mapping initiatives, some slum dwellers actively resisted data capture, believing that being under the ‘gaze of the state’ would have negative repercussions:

If given a choice (which many were not) most slum dwellers had expressed primacy of their right of representation: to be incorporated into urban datasets. But for some this was seen to be in tension with their right to privacy, and for a few who wished to ‘fly under the state’s radar’, the latter was dominant. (p. 1001)

Some businesses such as schools and pharmacies in Kibera did not wish to be mapped. They feared visibility to the state might lead to closure if their location became known and their informal status or activities (e.g., sales of stolen drugs) were then discovered….Particular settlements in Chennai refused to participate in data-gathering. They believed that drawing attention to their existence and informal status…would increase likelihood of eviction….Transparent Chenna itself had concerns about this. For example, it captured data on issues facing informal waste-pickers in the city but not their location or legal status, in order to protect them from state action. (p. 1001)

One of the major strengths of this framework lies in its utility to offer a comprehensive, retrospective evaluation of the data justice issues arising from the introduction of a new data system. The examples of community mapping discussed by Heeks and Shekhar involve datafication, but not algorithmic intervention. However, we do not have to look too far to find such an instance and one with many issues of data justice.

Automated Neglect

Human Rights Watch (2023) recently published a report titled Automated Neglect: How The World Bank’s Push to Allocate Cash Assistance Using Algorithms Threatens Rights. In it, they outline the many breaches of human rights associated with a World Bank sponsored programme in Jordan called the Unified Cash Transfer Program, although commonly referred to as, Takaful (an Arabic term sometimes translated as solidarity or mutual cooperation). As the report states:

After screening out families that do not meet basic eligibility criteria, Takaful uses an algorithm to identify which of those remaining should receive cash transfers by ranking their level of economic vulnerability…Human Rights Watch found that this algorithm is leading to cash transfer decisions that deprive people of their rights to social security. The problem is not merely that the algorithm relies on inaccurate and unreliable data about people’s finances. Its formula also flattens the economic complexity of people’s lives into a crude ranking that pits one household against another, fueling social tension and perceptions of unfairness. (p1)

Reading through the report, it is not difficult to map the issues identified by Human Rights Watch to all five of the dimensions of the conceptual framework. For example, the following illustration suggests issues with both procedural and instrumental dimensions of data justice:

Some people told Human Rights Watch that owning a car could have been one of the reasons they were rejected from Takaful, even though they needed it for work, or to transport water and firewood. “The car destroyed us,” said Mariam, a resident of al-Burbaita village in the southern governorate of Tafilah, one of the poorest villages in the country. Her family received Takaful cash transfers in 2021 but was dropped from the program in 2022. “We use it to transport water and for other needs. But sometimes we don’t have the money to fill it up with diesel, so we walk to the street and wait for someone to pass by and agree to pick us up,” she added. (p.4)

Using the framework

Although the primary purpose of the Heeks and Shekhar framework is to evaluate international development datafication projects retrospectively, it can also be used prospectively to sensitise the designers of data-driven initiatives to the principles of data justice and ensure they are designed into initiatives from the outset. Clearly, this did not occur in the case of Takaful.

In addition to the detailed conceptual framework for data justice, Heeks (2017) has also developed some general data justice principles in a ‘Data-Justice-for-Development Manifesto’:

1. Demand just and legal uses of development data.
2. Demand data consent of citizens that is truly informed.
3. Build upstream and downstream data-related capabilities among those who lack them in developing countries.
4. Promote rights of data access, data privacy, data ownership and data representation.
5. Promote data system outcomes that address international development goals and priorities; including the goals and priorities of data subjects.
6. Support ‘small data’ uses by individuals and communities in developing countries.
7. Advocate sustainable use of data and data systems.
8. Create a social movement for the ‘data subalterns’ of the global South.
9. Stimulate an alternative discourse around data-intensive development that places issues of justice at its heart.
10. Develop new organisational forms such as data-intensive development cooperatives.
11. Lobby for new data justice-based laws and policies in developing countries (including action on data monopolies).
12. Open up, challenge and provide alternatives to the data-related technical structures (code, algorithms, standards, etc.) that increasingly control international development.

An Applied Data Justice Framework for Datafication and Development

Other perspectives

• Framing data justice: A global perspective
• Framing data justice: A social justice perspective